Skip to content

Four different buck circuits on a single PCB benchmarked

License

Notifications You must be signed in to change notification settings

TimGoll/pcb_buck_test

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

10 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Buck Test PCB

To get an understanding of different buck ICs and their drawbacks and benefits I designed this PCB. While all of them worked, some performed better than others. To verify my first impressions while initial testing I benchmarked the ICs over different voltages and currents.

The testing PCB (the four different buck circuits that I tested)

ICs I tested

  • XL1509-ADJE1 [datasheet]
    • input voltage range: 4.5..40V
    • output voltage: 1.23..37V (fixed versions available too)
    • output current: 2A
    • switching frequency: 50kHz/150kHz
  • MT2492 [datasheet]
    • input voltage range: 4.5..16V
    • output voltage: 0.8..15V
    • output current: 2A
    • switching frequency: 600kHz
  • TPS62160DSG [datasheet]
    • input voltage range: 3..17V
    • output voltage: 0.9..6V (fixed versions available too)
    • output current: 1A
    • switching frequency: 2.25MHz
    • can be used to create negative voltage

Testing Results

XL1509-ADJE1

XL1509-ADJE1 voltage range (output voltage vs different output currents)

XL1509-ADJE1 efficiency range (efficiency vs different output currents)

While this IC is very cheap and has a wide range of input voltages, everything else about it isn't worth it. The low switching frequency is the reason for the huge passive components that have to be added around this IC. This negates the benefit of its price since the required components are rather expensive.

Moreover this IC really struggled with thermals. I had to stop the tests at 1750mA due to the output becoming instable because the IC overheated.

MT2492

MT2492 voltage range (output voltage vs different output currents)

MT2492 efficiency range (efficiency vs different output currents)

The MT2492 is a cheap and less capable copy of the MP2492 by MPS. It comes in a small package that combined with the smaller passive components needs very little space on the PCB.

Its efficiency is rather good and the output voltage is stable. Even over the time of multiple minutes the IC did not overheat when 2A were pulled.

All in all this is a good IC when more than 1A of continuous current is needed.

TPS62160DSG

TPS62160DSG voltage range (output voltage vs different output currents)

TPS62160DSG efficiency range (efficiency vs different output currents)

This IC made by TI is the smalles of them all. Paired with its high switching frequency the overall footprint of the circuit is really small. The output voltage is very stable, even for small drops and rather high currents. The efficiency is very good as well.

This is the best of the bunch, even though it is a bit more expensive. The most impressive thing is that it did not overheat, even though my thermal design for this circuit was very bad.

TPS62160DSG (Inverted)

TPS62160DSG voltage range (output voltage vs different output currents)

TPS62160DSG efficiency range (efficiency vs different output currents)

The same IC from the previous test was used again, but in a special configuration. With the addition of a few further passive components this IC can be used as an inverted power supply. While the input voltage range is decreased that way, the overall performance is still great.

One thing to highlight that it is possible to get stable -5V from a 5V source for up to 450mA of continous current. This makes this IC great for the use in OpAmp circuits where a negative rail is needed.

A minor drawback is that the efficiency is greatly reduced.

About

Four different buck circuits on a single PCB benchmarked

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages